Board Interlocking In Brazil: Directors’ Participation In Multiple Companies And Its Effect On Firm Value

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1

Board Interlocking in Brazil: Directors’ participation in multiple companies and its effect on firm value

Rafael Liza Santos

a

School of Economics, Management and Accounting, University of São Paulo (FEA/USP),

Brazil

Alexandre Di Miceli da Silveira

b

School of Economics, Management and Accounting, University of São Paulo (FEA/USP),

Brazil

Lucas Ayres B. de C. Barros

c

Mackenzie Presbyterian University and School of Economics, Management and Accounting,

University of São Paulo (FEA/USP), Brazil a Graduate Student at School of Economics, Management and Accounting of the University of São

Paulo (FEA/USP). +55 11 7215-2238. e-mail: rafael.liza@usp.br

b Professor of Finance and Accounting at School of Economics, Management and Accounting of the

University of São Paulo (FEA/USP). +55 11 5054-1888. e-mail: alexfea@usp.br

(contact author). c Professor of Finance at Mackenzie Presbyterian University, Professor of Finance and Accounting at

School of Economics, Management and Accounting of the University of São Paulo (FEA/USP). +55

11 3871-2689. e-mail: lucasayres@mackenzie.br

© Rafael Liza Santos, Alexandre Di Miceli da Silveira & Lucas Ayres Barros, 2008. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full credit, including © notice, is given to the source.

June 24-26, 2009

St. Hugh’s College, Oxford University, Oxford, UK

1


Board Interlocking in Brazil: Directors’ participation in multiple companies and its effect on firm value

Abstract

This paper investigates the simultaneous participation of directors in different companies from a sample of 320 Brazilian listed firms in 2001, 2003 and 2005. We identify which firms are connected through a network of directors, which corporate characteristics contribute to this phenomenon, and if board interlocking influences firm value and profitability. Our results show that interlocking directorates is a common practice in Brazil among listed firms.

Besides, larger boards, more dispersed ownership structures, and larger firm size are associated with higher levels of board interlocking. We also find that firm value is, on average, negatively impacted by interlocking, especially in firms with a “busy board” (when a majority of directors hold three or more directorships) or in firms where the CEO holds directorships in other companies. Overall, this paper provides support for the elaboration of governance guidelines by regulators regarding the consequences of interlocked directorates for corporate value.

Key-words : board interlocking, board of directors, corporate governance, directors, firm value

JEL classification code:

G32, G34.

June 24-26, 2009


2


1. Introduction

The board of directors is one of the main internal corporate governance mechanisms due to its roles of monitoring managers, ratifying relevant decisions, and setting the company’s strategic guidelines. Virtually all corporate governance codes of best practices highlight the importance of independent board members and the separation between CEO and chairman positions in order to ensure unbiased and impartial decisions (e.g. OECD, 1998).

However, one aspect that has received little attention in the literature refers to the possibility of corporate value being affected by the number of board seats occupied by the directors of the companies. This phenomenon is known as board interlocking

1

. It happens when a director serves as a board member in two or more companies, thus establishing a connection between them (Fich and White, 2001). Examples of research in this area can be found in Dooley (1969), Allen (1974), Mizruchi (1996), Fich and White (2001), and Fich and

Shivdasani (2006), among others. In Latin America, however, the subject remains unexplored and, to our knowledge, no studies have yet been published in this research line.

There are different arguments for the existence of this phenomenon: behavioral economists suggest that companies use board interlocking as a mechanism to improve the contractual relations between different firms, thus reducing implicit uncertainties (Schoorman,

Bazerman and Atkin, 1981). Other researchers argue that the phenomenon is simply driven by market competition, since better trained directors constitute a scarce resource like any other, so that companies will compete for the presence of these professionals on their boards 2 (Fich and White, 2001, Mizruchi 1996).

Regarding the effect of board interlocking on company performance, Mol (2001) argues that linking companies through board members may bring benefits resulting in competitive advantages, such as easing the access to resources, clients and creditors, disseminating innovations and effective methods, and positively influencing the environment

June 24-26, 2009


3

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 in which these companies are inserted. On the other hand, Fich and Shivdasani (2006) report that an analysis of the 500 largest American companies between 1989 and 1995 indicates that the presence of directors in multiple boards produces a negative effect on firm’s market value, supposedly caused by the deterioration of governance quality 3 .

Our study aims to understand and describe the occurrence of directors’ simultaneous participation in multiple companies in Brazil. We examine the board composition of all actively traded companies in the São Paulo Stock Exchange (Bovespa) in 2001, 2003, and

2005 .

We also investigate which corporate characteristics favor the occurrence of a higher degree of board interlocking in Brazil. Finally, employing different econometric procedures, we evaluate the impact of this practice on corporate value and profitability.

We provide three main results: firstly, we observe that board interlocking is a recurrent practice in Brazil among exchange listed firms. In 2003, for instance, the 3,196 board seats of the firms under analysis were occupied by 2,688 individuals. Besides, 74% of the sampled firms had at least one officer or director simultaneously working for another company. We also find that the composition and size of the board, firm size, and the degree of stock ownership dispersion are the main attributes that seem to influence the extension of this phenomenon. Finally, in line with other studies in this field, our analysis suggests that board interlocking tends to negatively impact firm value. In fact, two types of board interlocking seem to be particularly damaging to corporate value: the participation of the CEO and/or the chairman of the board in other companies’ boards, and the presence of a large ratio of outside directors participating in different companies (the so-called “busy boards” 4 ).

Our motivation for this research is based on the argument that the interconnections among companies through their directors is a mechanism that can affect corporate strategies and decisions that, at the macroeconomic level, can increase or decrease overall market efficiency. As an illustration of such implications, Bizjak et al. (2007) provided evidence that

June 24-26, 2009


4

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 recent corporate governance problems in the U.S. market, caused by option backdating in order to raise executives’ gains, may have been disseminated by “word-of-mouth” among directors simultaneously serving in different boards.

Overall, we believe that this research offers three mains contributions: i) it is a pioneering exploratory research on this field in Brazil and, to our knowledge, in Latin

America, therefore contributing to the international literature by providing empirical evidence from an emerging country; ii) we offer evidence on the determinants of board interlocking, showing which corporate characteristics tend to influence the adoption of this practice to a greater or lesser extent; and, iii) we provide evidence, using different proxies for board interlocking, that this practice may have a negative influence on firm value. This result is particularly important, since it offers empirical support for regulations or recommendations on this subject in codes of best practices of corporate governance.

The paper is organized as follows: Section II presents a literature review on the subject and discusses the research hypotheses; Section III addresses the methodological and sampling issues; Section IV presents an overview of board interlocking in Brazil, through descriptive statistics and an example of mapping interlocked companies; Section V discusses which factors favor the formation of board interlocking based on results from multivariate probit regressions; Section VI assesses, using alternative econometric procedures, the extent to which corporate value and profitability are affected by the board interlocking phenomenon; finally, Section VII concludes.

2. Literature Review and Hypothesis

2.1. Outside Directors and Board of Directors’ Effectiveness

In order to strengthen the monitoring role attributed to the board, corporate governance codes worldwide recommend the inclusion of independent outside directors on

June 24-26, 2009


5

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 company boards, that is, professionals who neither participate in nor are connected with the management team, as a way of improving decision making and, thus, raising company value.

The actual effectiveness of outside directors, however, is a controversial matter, as evidenced by the inconclusive results linking corporate value to board independence.

It is among outside directors that board interlocking is most frequent

5

, possibly because inside directors, working full time as officers, do not have available time to work in other companies. Hence, one of the main critiques against a director working in multiple companies is the possible lack of time for an adequate dedication to defend the interests of the companies’ shareholders 6

. In an event study, Fich (2005) reports a stock price decline for companies whose directors announce their participation in other boards, suggesting that shareholders fear that their interests will be damaged when board interlocking occurs 7 .

Thus, it seems reasonable to presume that outside directors’ effectiveness is conditioned not only by their independence from the CEO, but also by their degree of dedication and commitment: members who already participate in other companies (mainly those participating on many boards), despite having more contacts or experience, may not perform their task as effectively as other directors. Hence, board interlocking appears in the corporate governance literature as a potential determinant of board effectiveness.

2.2. Board Interlocking: definitions and potential determinants

According to Mizruchi (1996), “the board interlocking occurs when a person affiliated with one organization sits on the board of directors of another organization”. The investigation of such firm linkages has aroused interest in different research areas, such as management, economics, sociology and law. Mizruchi (1996) underlines that board interlocking has become the primary indicator of connection among firms. Therefore, its analysis could provide clues about important aspects of the behavior of firms.

June 24-26, 2009


6


Two hypotheses stand out as justification for board interlocking: the first argues that companies benefit when their board members bring better contacts – such as clients, suppliers, creditors, new experiences – or when directors monitor connected firms. As a result, companies would stimulate their directors to participate in other companies or only would like to hire people serving on other boards. A second hypothesis argues that the directors’ desire to participate in other boards is the sole responsible for the occurrence of board interlocking, independently of the interests of the companies.

Works such as Pennings (1980) and Burt (1983) offer support to the first hypothesis .

The authors report that, in industries where few companies concentrate most of the market share, there are more interlocked boards. Thus, board interlocking could work as a facilitator for collusion among competitors, favoring oligopolistic market practices. In addition,

Carrington (1981) finds a positive relation between market concentration, interconnections between companies and profitability. Overall, Mizruchi (1996) argues that, although access to empirical evidence is difficult, intra-industry board interlocking can be a way of imposing restrictions on market competition.

Some studies also interpret board interlocking as a cooptation and monitoring mechanism by firms. In this view, it could act as an instrument to reduce sources of uncertainty in the economic environment because shared directors could improve the information flow to their companies. In a similar way, monitoring a certain company by including professionals from another firm in its board is another common form of cooptation.

A practical example frequently mentioned in this literature refers to highly financially leveraged companies: they usually take their creditors’ representatives in their board of directors, in order to reduce the risk of default (e.g.: Dooley, 1969; Pfeffer, 1972).

Under this perspective of company interdependence, Nicholson et al. (2004) argue that ownership dispersion is positively associated with board interlocking, since the board of

June 24-26, 2009


7

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 companies with a high degree of ownership dispersion would include members from different origins and with different relations with the company: creditors, clients, suppliers and professional directors. Similarly, the concentration of control in family-held companies would be negatively correlated with their directors’ participation in other companies (Allen, 1976;

Burt, 1980). Finally, another argument contends that the board of directors plays an important role in protecting the firm’s reputation with its investors (Selznick, 1957; Parsons, 1960).

Thus, the fact that a director from an important company i is invited to serve on the board of another firm j may indicate to investors that company j is trustworthy.

The second view on board interlocking rationale states that it is not only because of their company’s interests that professionals participate in more than one board, but especially because of their self interest. Therefore, board interlocking could be an involuntary consequence of these professionals’ individual decisions instead of a result of corporate strategy. Palmer (1983) corroborates this view, by showing that interlocked companies did not reconstitute the links that were accidentally broken (e.g. due to the death or retirement of a director). Zajac (1988) and Stokman et al. (1988) also observe that directors participate in boards because of financial remuneration, prestige and contacts that can be useful for future career advances.

When analyzing reciprocal board interlocking among CEO’s of the 500 largest

American companies, Fich and White (2001) observe that the probability of occurrence of this phenomenon decreases as the number of stock options in the CEO’s compensation package increases, which leads to the conclusion that, the more their remuneration depends on company performance, the more professionals become dedicated to their tasks in the

‘original’ firm. Dooley (1969) and Pfeffer (1972) also find that the degree of share ownership by managers is negatively correlated with the formation of board interlocking.

June 24-26, 2009


8


Fich and White (2001) note that the greater the number of annual meetings of the board, the lesser the probability that their members will participate in other firms, a fact justified by the unavailability of time for the directors’ activities. They also find that bigger boards are more likely to have a member from another company and that professionals with longer company time develop more contacts and, therefore, tend to be invited to participate in other companies.

Although we cannot empirically test all of the arguments discussed above for the

Brazilian market, especially due to the lack of data on managers’ remuneration or the frequency of board activities, we are able to test the following hypotheses, based on this literature review:

H

1

: The probability that board interlocking will occur in a company increases with the number of members in its board of directors.

H

2

: A greater participation of outside directors tends to increase the probability that board interlocking will occur.

H

3

: The occurrence of board interlocking is more likely in more renowned

8 companies.

H

4

: The degree of concentration of stock ownership is inversely related to the probability of occurrence of board interlocking.

H

5

: Financial leverage positively affects the probability of occurrence of board interlocking.

We address below the potential effects of board interlocking on firm profitability and value, and derive from the literature other testable hypotheses.

2.3 Consequences of Board Interlocking for Corporate Performance and Value

Although many empirical studies on the subject show a negative correlation between directors’ participation in multiple companies and firms’ financial indicators (Fich and

June 24-26, 2009


9


Shivdasani, 2006; Fich and White, 2001; Loderer and Peyer, 2002), there are some theoretical studies predicting beneficial results associated with this practice.

Research in organizational sociology argues that the board of directors holds a very valuable social capital for the company, due to two main reasons: their contacts and their knowledge. Nicholson, Alexander and Kiel (2004) highlight that board interlocking is especially advantageous for companies facing scenarios of uncertainty and interdependence in the market, organizational complexity and dispersed ownership, since this mechanism would act against possible disadvantages of these features through the directors’ social connections and their influence in the market environment.

The concept of social cohesion, according to Mizruchi (1990), is associated with members sharing similar interests and with an extensive network of social contacts (allowing a large flow of relevant information). Through these factors, top management professionals are inserted in a club-like social environment: dense, cohesive and private, which becomes active and potentially useful to its participants through their social contacts.

Using the Transaction Cost Economics framework

9

, the authors theorize that companies whose directors share top management functions would have more communication routes and better information exchange, which could reduce transaction costs. Boyd (1990) finds that, among firms who face greater uncertainty in the economic environment, those with greater connections with other companies through shared directors have better performance – measured by sales’ increase and return on equity.

Another mechanism through which boards with connections in other companies add value to the company refers to the cognitive capital these members acquire, i.e., their personal experience

10

. Haunschild and Beckman (1998) note that connections between directors allow the dissemination of organizational innovations through this corporate contact network. The similarity of strategies adopted by interconnected companies is addressed in many studies.

June 24-26, 2009


10


Palmer, Jennings and Zhou (1989) report that American companies showed stronger tendencies to adopt a multidivisional structure if they had interlocks with companies that had already adopted this structure. Davis (1991) shows that the adoption of poison pills is faster in companies connected with firms that already had this anti-takeover mechanism .

D’Aveni and

Kesner (1993) observe that, in companies whose top management shared multiple directories, hostile takeover attempts were more successful in comparison with target companies that did not have any connection. Even similar political campaign contribution strategies are identified in companies connected through their directors. When analyzing a sample of 243 American connected companies among the 250 largest in that country, Clawson and Neustadtl (1989) observe that the firms linked by their directors were more inclined to contribute to the same political group during the American elections in 1980.

Most sociological-organizational studies on this theme are theoretical. Despite these arguments based on social cohesion and cognitive capital accumulation by professionals who share multiple directorates, we know of no empirical study that shows long-term economic benefits of interlocked boards.

On the other hand, some research in corporate finance and economic behavior offer more quantitative work, trying to uncover the effects of board interlocking on the financial indicators of companies. Most studies find a negative correlation between the frequency of directors’ participation in more than one firm and these companies’ financial results. Fligstein and Brantley (1992) identify a negative association between interlocks and profitability, in a sample of large American corporations. In an analysis of companies listed in the Swiss stock exchange (Zurich Stock Exchange) between 1980 and 1995, Loderer and Peyer (2002) show that the accumulation of functions in other boards reduces firm value, measured by Tobin’s

Q

11

. The authors justify that the possible reasons for these results are the conflicts of interest

June 24-26, 2009


11

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 directors are exposed to by participating in multiple companies – which would increase the potential for agency costs – and the lack of adequate time to perform multiple mandates

12

.

In an analysis of 1,493 indications for outside directors, Fich (2005) observes significantly positive reactions of share prices when CEOs from other companies were indicated as outside directors. This result suggests that the market perceives the CEOs of other

American companies as having knowledge and contacts especially useful to their company.

On the other hand, the author also reports that these CEOs’ original firms suffered a share price decrease upon their indication to serve as outside board members in other companies, which might be indicative of the ambiguous nature of board interlocking: beneficial for firms that acquire an experienced outside director, but harmful to the ‘original’ firm, whose CEO will be busier than before and, hence, may dedicate less attention to maximizing her company’s shareholder return.

When analyzing reciprocal board interlocking among CEOs in a sample of 610 of the largest American corporations, Fich and White (2001) identify a positive association between

“two-way” participation of CEOs and the Tobin’s Q of the companies that exchanged these professionals 13 .

Opposing the argument that a larger number of contacts, through the company’s directors, results in better opportunities and, consequently, in better corporate performance,

Fich and Shivdasani (2006) develop the concept of “busy boards”, defined as the condition in which half or more of the company’s outside directors serve on three or more boards. Under this perspective, the authors avoid the distortion of an approach that uses the mean number of seats in other firms per board member, in which case one single director participating in many boards would bias the mean interlocking of the company. Using this qualitative measure of busy boards, the authors show that board members affiliated with too many other companies reduce corporate value and performance. The authors verify that the price-to-book ratio of

June 24-26, 2009


12

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 companies whose boards are classified as busy are 4.2% lower than that of companies whose board does not fit into this classification.

These results lead us to the hypothesis that the marginal benefits of the better contacts and greater experience of directors who participate in multiple companies do not exceed the decrease in the quality of their monitoring activities, so that board interlocking is generally negative to the company. This argument is formally stated below.

H

6

: A higher level of board interlocking is associated with lower company value, ceteris paribus. This impact should be more negative in companies with a greater proportion of board members acting on multiple boards (called “busy boards” in this research).

3. Sample, Variables and Database

Our sample consists of all companies listed on the São Paulo Stock Exchange

(BOVESPA) whose liquidity index

14

was higher than zero in 2001, 2003 and 2005. The gap of one year between the observations is not a problem for our analysis because there is little variation observed in the degree of board interlocking, since the mean and median term of the directors’ tenure is approximately two years. After the selection of the sample, with 345 companies for 2001, 319 for 2003 and 320 for 2005, we collected data from all board memberships in the Annual Reports (AR) submitted to the Brazilian Securities and Exchange

Commission (CVM). Overall, we have 4,460 top management observations constructed for

2001, 3,392 for 2003 and 3,156 for 2005, totaling 11,008 directorships.

Next, we assess for each director how many times his/her name

15

appeared in the sample. We identify which professionals occupied more than one seat and in which company this seat was occupied, thus allowing the mapping of interconnected companies. However, as a limitation of our analysis, we are not able to assess the participation of directors in privately

June 24-26, 2009


13

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 held companies, since they do not disclose their board composition. This can bias our data, underestimating the real degree of board interlocking.

For each company, we assess the size of the board of directors, as well as the proportion of inside and outside directors 16 , and the total number of companies with at least one shared board member. Other binary variables were computed, indicating if the company was connected to another company in the sample through its board members, if the CEO or chairman of the board occupied a seat in another company, and if more than half of the company’s outside directors worked in three or more companies (‘busy board’ variable).

For the ownership structure, we c the percentage of voting stocks held by the controlling block

17

identified in the ARs. The companies listed in the two top special listing corporate governance segments (Level 2 and New Market) were identified through the

Bovespa’s website 18

.

Financial data, such as the market value of stocks, total assets, net equity, gross debt, operational income, etc., were collected in the Economática® database system for each of the four quarters of the three years analyzed. We also identify to which industry the company belongs and whether it issues American Depositary Receipts (ADRs) Levels 2 or 3, therefore being subject to the Sarbanes-Oxley requirements.

All firms were considered for the calculations of the descriptive statistics. However, the tests aiming to evaluate the influence of board interlocking on firm value were restricted to a sub-sample of 152 non-financial firms with a liquidity index higher than 0.05% of the most liquid company in each year, during the 3 years under analysis, resulting in a balanced panel with 456 firm-year observations. Our results are presented below.

4. Overview of board interlocking in Brazil: Descriptive Statistics

4.1. The Board of Directors

June 24-26, 2009


14


Table 1 shows descriptive statistics regarding the board composition of the sampled firms.

[Please insert table 1 here]

According to Table 1, the mean board size is 6.45 for the overall sample. We also note that most of the directors do not work as officers. Therefore, our proxy for the level of independence of the board, computed as the percentage of non-executive board members, has a mean of 81%. It is important to note, however, that being an outside director does not mean that he/she is automatically independent.

We also find that in more than 30% of the companies, the same person holds both the

CEO and chairman positions (this is reported In Table 2). This practice could hinder the boards’ supervisory function, possibly increasing agency costs deriving from bad monitoring of top managers.

4.2. Board interlocking

Table 2 describes the patterns of board interlocking in our data.

[Please insert Table 2 here]

According to Table 2, connections of firms via shared directors are not unusual in

Brazil. In 2005, more than 68% of the sampled companies were interlocked with at least one board member serving as a director of another company listed in the stock exchange

19

.

Another questionable finding in terms of governance practices is the fact that about

35% of the companies’ CEOs held a top management position in another company 20

.

June 24-26, 2009


15


Likewise, in 2005, about 50% of the chairmen of the board served as directors in other companies. These observations suggest some cohesion in the business class in ruling the biggest companies in Brazil. Therefore, there is evidence of the existence of a vast network of social contacts and shared interests, allowing the accumulation of positions by these professionals.

About 40% of the firms have a board in which half or more than half of their outside directors participate in at least one other company. These companies could face substantial agency problems because their directors might be too busy with other activities. Likewise, more than a quarter of the companies have the so-called “busy boards” – when half or more than half of the outside directors participate in three or more companies, including their

“original” firm. This situation casts doubt on these professionals’ effectiveness as monitors.

Table 3 provides information on the number of interlocks at the firm level.


In Table 3, we note that, in 2001, the mean number of interlocks with other companies was 3, against 5.5 in 2003 and 4.8 in 2005. Moreover, among a subsample of 150 most liquid companies in each year, 85 companies were interlocked in 2001, while 96 companies have at least one director from another firm in 2005.

Another interesting perspective can be provided by mapping the connections between companies. Chart 1 illustrates, in a simplified manner, one of the networks found in 2005 and shows that the network of interlocks among firms can be dense and pervasive.

[Please insert chart 1 here]

June 24-26, 2009


16


4.3. The directors and their directorships

Table 4 presents evidence on the frequency of board interlocking in Brazil.


According to Table 4, the directors and officers in our sample held, on average, 1.26 top management positions in 2003. That is, for every 5 top management positions, there are only 4 professionals to hold them, on average. Although the majority of the professionals are exclusively dedicated to their original company, we have found one board member simultaneously participating in eight different companies in 2003.

5. The determinants of board interlocking in Brazil

To test the hypotheses H1 to H5 discussed above, we use a multivariate probit regression to model the (conditional) probability of the occurrence of board interlocking, as shown below.

Pr( INTERLOCK it



1 X it

)

  













5

 

1

BDSIZE

DEBT it

 

6 it

 

2

INDEP

TOBINSQ it

 it



7

 

3

CEOCHAIR

ROA it

 

8 ln( it

 

ASSET

4 it

CONTROL it

)

 

9

BOV it



 

10

YEAR t

 i

20 



1

 j

IND it











Equation 1

Where:

INTERLOCK it



1 means that company i in year t had at least one board member from another company



INTERLOCK it



0 otherwise



; X represents the set of regressors shown in the right hand side, as follows: BDSIZE (Board size); INDEP (Independence, proxied by the percentage of outside directors in the board); CEOCHAIR (binary variable, indicating if the

CEO and chairman positions are held by the same person); CONTROL (percentage of voting shares held by the controlling shareholders); DEBT (natural logarithm of the ratio of gross debt to total assets, i.e., financial leverage); TOBINSQ (Tobin’s Q

21

); ROA (Return on assets –

June 24-26, 2009


17

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 operational income divided by total assets); ln( ASSET ): (natural logarithm of mean assets;

BOV (binary variable indicating if the company is listed in Bovespa’s Level 2 or New

Market); IND (industry binary variables); YEAR (year binary variable).





denotes the cumulative normal distribution.

The main results of the regression are presented in Table 5.


As shown in Table 5 and discussed below, our results support all but one of the hypotheses H1 to H5 : i.

There is a positive influence of board size on the likelihood of board interlocking, that is, boards with a larger number of members tend to accept the participation of board members from other companies or, from another point of view, tend to send their directors to other companies, thus creating links among them. ii.

Board independence is positively associated with board interlocking: participation of outsiders on the board favors the formation of interlocks. iii.

There is a negative influence of ownership concentration on board interlocking. This result may be explained by the fact that firms with a higher share concentration have smaller boards, with a higher proportion of directors dedicated to the interests of the controlling group, so these professionals usually do not serve in other companies. Moreover, the concentration of control tends to reduce outsiders’ participation in the board. iv.

Firm size shows a positive and significant influence on the likelihood of board interlocking, i.e., larger companies tend to have more board members from other companies. v.

Financial leverage does not significantly influence the formation of board interlocking.

June 24-26, 2009


18

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 vi.

Finally, our results support the hypothesis that firms with better reputation have a higher probability of board interlocking. We use a dummy indicating that the company joined the differentiated corporate governance levels of Bovespa as a proxy for reputation and that variable is positively and significantly related to board interlocking. The idea is that these firms may have directors with better than average reputation in the corporate management market, who would therefore be invited to serve in other companies.

6. Board interlocking and corporate value

In order to assess the effect of board interlocking on firm value, we run the following model with four different estimation methods: random effects, fixed effects, two stages least squares (2SLS) and the generalized method of moments (GMM).

V it

 α  β

1

INT it

 β

2

BSIZE it

 β

3

INDEP it

 β

4

CEOCHAIR it

 β

5

CONTROL it



 β

6 ln( ASSETS it

)

 β

7

ROA it

 β

8 ln( DEBT it

)

 β

9

YEAR t

 i

20 



1

 j

IND it

 ε it

Equation 2

For robustness purposes, we test two different proxies for corporate value ( V

): Tobin’s

Q and Price-to-Book Value ratio ( PBV ). We also use four alternative measures of interlocking

( INT ): A dummy for the existence of this phenomenon; a dummy for the participation of CEO or chairman of the board in other firms; the total number of board interlocks of the firm, and a dummy variable called ‘busy board’, following the definition by Fich and Shivdasani (2006), when half or more of the outside board members participate in 3 or more companies. All tests indicate that board interlocking reduces corporate value.

6.1 Fixed Effects and Random Effects for Panel Data

Fixed Effects estimation allows us to isolate unobserved firm characteristics that could affect simultaneously the frequency of board interlocking and our dependent variables, thus

June 24-26, 2009


19

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 producing more reliable coefficients than the ones estimated with OLS. These results are shown in Table 6.


In Table 6, the estimates for the interlocking proxies all have negative sign but are not statistically significant, except for model number 4, in which we use the ‘busy board’ concept as the proxy for the degree of interlocking. It is possible that the lack of significance in the first three models is caused by the little variation in the degree of interlocking over time. This is the case because boards of directors generally have a stable composition in a 2 or 3 years time span. Because of that, the fixed effects estimator – which isolates time invariant firm characteristics – may not be the most adequate to assess the board interlocking effect on firm value.

Alternatively, we estimate the same models using the random effects procedure.

Again, we note that all interlocking proxies have negative coefficients. The most significant ones were the dummy for participation of CEO or Chairman in other company and the busy board dummy, both statistically significant at the 5% level.

6.2 Two Stage Least Squares Estimations (2SLS) and GMM Estimations

Our random and fixed effects estimations do not deal with some endogeneity issues, especially the problem of simultaneity. However, as discussed in the extant board interlocking literature, this is probably a relevant concern: interlocked directorates may reduce firm value but it is also possible that companies who already have a poor performance induce their directors to look for a position in other companies. Therefore, the interlocking can be both a

June 24-26, 2009


20

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 cause and a consequence of poor firm performance. To tackle this endogeneity problem, we use an instrumental variable approach.

In line with previous works (e.g. Fich e Shivdasani, 2006), we run two stage least squares regressions using one period (two years) lagged values of our interlocking proxies as instruments. These instrumental variables are correlated with the (contemporaneous) interlocking proxies and address the simultaneity issue. The results are shown in Table 7.


According to Table 7, all estimates are negative for the interlocking proxies and the coefficients are generally more significant. In sum, they confirm the hypothesis that higher levels of board interlocking are harmful to the company (or to its shareholders). We obtain virtually the same results when using the Tobin’s Q or PBV as proxies for firm value. We also observe a negative and strong effect on firm value when the CEO and Chairman positions are held by the same person. This result goes in line with the literature, which points that the monitoring effect of directors is reduced when the CEO is also the Chairman of the board.

We obtain similar results when using a GMM estimator and lagged values of our interlocking proxies as instruments. Thus, our data seems to reject the hypothesis that company value would benefit from a larger number of interlocks. On the opposite, our analysis suggests that a larger number of participations by directors in other companies is not a positive factor for company value. For the third model, where we estimate the effect of the number of company ties associated with shared board members, we obtain a negative coefficient, indicating that, at each new linkage established by the directors, company value would decrease about 5%, ceteris paribus .

June 24-26, 2009


21


It is possible, as Loderer and Peyer (2002) suggest, that the professionals who participate in these networks are actually involved with too many activities, making them too busy to be good monitors of the firm, which would lead to a deterioration of theirs firm’s corporate governance quality.

We use, in model 4, the ‘busy board’ concept and note that, when at least half of the outside directors participate in 3 or more companies, firm value, measured by its Tobin’s Q, is considerably reduced. From an economic perspective, considering that the mean Tobin’s Q in the sample is 1.29, this finding would mean a decrease in company value by about 28%. We find very similar results when using price-to-book value as the dependent variable.

We also found significant coefficients for other explanatory variables. The concentration of ownership, for instance, seems to negatively influence company value, which is indicative of an agency problem between controlling and minority shareholders in Brazil.

6.3 Robustness Cheks

As discussed in the literature review, board interlocking might affect company performance due to four main reasons: 1) it is supposed to be a mechanism for manager cooptation and monitoring between connected companies, reducing uncertainties and transaction costs; 2) it could allow for more personal contacts and increase information flows through companies’ top management, providing valuable data for decision making and disseminating successful strategies; 3) interlocking could reflect a situation of social cohesion, in which only a restricted group of professionals would participate, a fact that would make the relations between these directors more personal and reduce their independent critical judgment; 4) interlocking could overload directors with tasks that would reduce their time and, hence, would be negative for the quality of their monitoring and decisions.

June 24-26, 2009


22


The first two hypotheses suggest a positive effect of interlocking for the firm, while the latter two predict harmful consequences of this practice. Therefore, it is not unreasonable to hypothesize that there may be a nonlinear relation between corporate performance and the degree of board interlocking. Specifically, we estimate quadratic models of the effect of board interlocking on firm performance. According to our hypothesis, when the level of interlocking is low, the directors are not too busy yet and their contacts could act in the benefit of the firm, so that a certain number of connections would be positive for the company. On the other hand, when the level of interlocking is excessively high, the board members would not have time to adequately perform their tasks and, moreover, adding more connections would not provide significant marginal benefits, resulting in a negative effect of excessive interlocking.

Therefore, as a robustness check, we add interlocking variables in their quadratic form to the previously estimated models. We test two alternative proxies separately: the natural logarithm of the number of connections per company and the percentage of external board members classified as busy in the board, i.e., those serving in 3 or more companies.

When using Tobin’s Q and PBV as dependent variables, the estimated coefficients of the interlocking proxies were generally non-significant. The exception was the regression of

PBV on the percentage of board members classified as busy, in which we found a positive quadratic relation

22

, which is counterintuitive. Overall, however, the instability of the estimates associated with these regressions does not warrant consistent inferences.

On the other hand, when we regress the board interlocking proxies on the profitability indicators return on assets (ROA) and return on equity (ROE), we find a negative (inverted Ushaped) and significant quadratic relation, in line with the intuitive abovementioned hypothesis about an optimal level of interlocking .

Specifically, an increase in the number of connections seems to positively influence the company’s ROA up to five connections, with a decrease in profitability from that point onwards. We find similar results for ROE, but with

June 24-26, 2009


23

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 lower statistical significance. In general, our results are similar to the ones reported by Non and Franses (2007), who analyzed 101 Dutch publicly-traded companies using a four-year time window

23

.

In sum, extreme levels of interlocking seem to negatively affect operational performance (and, thus, profitability), although the market did not reflect this relation on stock prices. These results are synthetically presented in Table 8 and in Figure 1, which depicts the distribution of ROA as a function of the number of firm interlocks.


[Please insert Figure 1 here]

7. Concluding remarks

Our study investigates whether board interlocking, the connection of firms by executives simultaneously serving in two or more boards, actually happens in Brazil and how often this phenomenon occurs. We also analyze which firms’ characteristics are more associated with this practice and the influence of this phenomenon on firm value and profitability.

We find that board interlocking is a recurring and pervasive practice among listed companies in Brazil. In 2003, 74% of the 319 sampled firms had at least one director from another company. In 2005 this ratio was 68%. We also found that the average professional held 1.26 top management positions within the sample. As a caveat, we should highlight that these figures probably underestimate the occurrence of interlocking, since we are not able to assess whether these individuals also serve as directors in private firms, which are not required to disclose information about their management team and board.

June 24-26, 2009


24


In line with previous related studies (Mizruchi, 1996, Fich and White, 2001), we find that companies with larger boards and boards with a higher ratio of outside directors tend to have more interlocks with other companies. Our analysis also suggests that companies with more assets and better reputation 24 tend to have more interlocks. On the other hand, companies with greater ownership concentration have less interlocked boards, possibly because of the influence of the control block on the selection of board members.

Regarding the influence of board interlocking on corporate value, our results clearly suggest that this practice negatively influences firm value. Specifically, the existence of a

CEO and/or chairman of the board serving in other companies tends to reduce corporate value, measured by its Tobin’s Q and price-to-book-value ratio. Another relevant finding is that busy boards, in which half or more external members serve in three or more companies, also negatively influence firm value. As a robustness check, we test the plausibility of a quadratic relation, attempting to capture the possibly nonlinear relation between interlocking and firm performance. Our regression analysis suggests that a quadratic relation is plausible when we use profitability measures as dependent variables, with an optimal interlocking of about 5 connections. Overall, our results suggest that the benefits of the interlocking, such as a better contact network of directors or their greater professional experience, do not compensate for the deterioration in these professionals’ quality of decision and monitoring capacity.

We argue, however, that these results are not evidence against the presence of professional independent directors on boards, which is regarded as essential among the socalled good corporate governance practices. In fact, our evidence should be understood as suggesting a negative impact of having many directors simultaneously participating in many boards, which might overload these professionals and, as a consequence, undermine their capacity to supervise managers and add value to their business, thus causing lower corporate

June 24-26, 2009


25

2009 Oxford Business & Economics Conference Program ISBN : 978-0-9742114-1-1 performance. In other words, the participation of independent outside board members may be positive and relevant, as long as they keep focused on the companies they serve in.

In sum, board interlocking seems to be a relevant phenomenon, particularly because of its potential negative influence on firm value. The analysis of this network of corporate interconnections also provides indication that the corporate class ruling the largest Brazilian companies is somewhat cohesive and interrelated. Therefore, this research might contribute to the understanding of the effectiveness of outside directors taking into account the accumulation of their professional duties as a relevant factor for the efficacy of their decisions. In particular, our study can subsidize recommendations about the maximum number of directorships for a director, taken as a relevant input for maximizing corporate value.

June 24-26, 2009


26


8. References

ALLEN, M.

The Structure of Interorganizational Elite Cooptation: Interlocking Corporate Directorates

(1974). The American Sociological Review , 39, 3, 393-406.

ALLEN, M. Management control in the large corporation: Comment on Zeitlin. (1976). American

Journal of Sociology , 81, 885-894.

BIZJAK, J. M., LEMMON, M. L., WHITBY, R. J. (2007). Option Backdating and Board Interlocks.

AFA 2008 New Orleans Meetings Paper . Available at SSRN http://ssrn.com/abstract=946787

BOYD, B. (1990) Corporate Linkages and Organizational Environment: A Test of the Resource

Dependence Model..

Strategic Management Journal , 11, 6, 419-430.

BURT, R. (1980). Cooptive Corporate Actor Networks: A Reconsideration of Interlocking Directorates

Involving American Manufacturing.

Administrative Science Quarterly , 25, 4, 557-582.

BURT, R. (1983). Corporate Profits and Cooptation , New York: Academic Paper.

CARRINGTON, P. (1981). Horizontal co-optation through corporate interlocks . PhD thesis,

Department of Sociology – University of Toronto.

CHUNG, K.; PRUITT, S. (1994). A simple approximation of Tobin’s Q. Financial Management , 23,

3, 70-74.

CLAWSON, D.

NEUSTADTL, A. (1989). Interlocks, PACs, and Corporate Conservatism.

The

American Journal of Sociology , 94, 4, 749-773.

CVM. (2002). Cartilha de Governança Corporativa . Available at http://www.cvm.gov.br

.

D’AVENI, R.; KESNER, F. (1993). Top managerial prestige, power, and tender offer response: a study of elite social networks and target firm cooperation during takeovers. Organization Science , 4, 2, 123-151.

DAVIS, G. (1991). Agents without principals? The spread of the poison pill through the intercorporate network. Administrative Science Quarterly , 36, 583-613.

DOOLEY, P. (1969). The Interlocking Directorate. The American Economic Review , 59, 3, 314-323.

FLIGSTEIN, N.; BRANTLEY, P. (1992). Bank Control, Owner Control, or Organizational Dynamics:

Who Controls the Large Modern Corporation?

The American Journal of Sociology , 98, 2, 280-307.

FICH, E. (2005). Are some outside directors better than others? Evidence from Director Appointments by Fortune 1000 firms. Journal of Business , 78, 5.

FICH, E.; SHIVDASANI, A. (2006). Are Busy Boards Effective Monitors? The Journal of Finance ,

61, 2, 689–724.

FICH, E.; WHITE, L. (2001). Why do CEOs reciprocally sit on each other's boards?. New York

University Center for Law and Business Working Paper #CLB-01-002 .

June 24-26, 2009


27


HAUNSCHILD, P.; BECKMAN, C. (1998). When Do Interlocks Matter? Alternate Sources of

Information and Interlock Influence. Administrative Science Quarterly , 43.

LIPTON, M; LORSCH, J. (1992). A modest proposal for improved corporate governance. Business

Lawyer , 48.

LODERER, C.; PEYER, U. (2002). Board overlap, seat accumulation and share prices. European

Financial Management , 8, 2, 165-192.

MIZRUCHI, M. (1996). Cohesion, Structural Equivalence, and Similarity of Behavior: An Approach to the Study of Corporate Political Power. Sociological Theory , 22, 271-298.

MIZRUCHI, M. (1990). What Do Interlocks Do? An Analysis, Critique, and Assessment of Research on Interlocking Directorates. Annual Review of Sociology , 8, 16-32.

MOL, M. (2001). Creating wealth through working with others: Interorganizational relationships. The

Academy of Management Executive , 15, 1, 150.

MORCK, R. SHLEIFER, A. VISHNY, R. (1988). Management ownership and market valuation: an empirical analysis. Journal of Financial Economics.

. Vol.20, p.293-315.

NICHOLSON, G.; ALEXANDER, M.; KIEL, G. (2004). Defining the social capital of the board of directors: an exploratory study. Journal of the Australian and New Zealand Academy of Management , 10, 1,

54-72.

NON, M., FRANSES, P. (2007). Interlocking Boards and Firm Performance: Evidence from a New

Panel Database. Tinbergen Discussion Paper .

OECD. (1988). OECD Economic Survey . Paris. Available at www.oecd.org.

PALMER, D. (1983). Broken ties; interlocking directorates and intercorporate coordination.

Administrative Science Quarterly , 28, 40-55.

PALMER, D.; JENNINGS, P.; ZHOU, X. (1989). Late adoption of the multidivisional form by large

U.S. corporations: institutional, political, and economic accounts. Administrative Science Quarterly , 38, 100-

131.

PARSONS, T. (1960). Structure and Process in Modern Societies, Glencoe, IL: Free.

PENNINGS, J. (1980). Interlocking Directorates: Origins and Consequences of Connections

Among Organizations' Boards of Directors , Jossey-Bass, San Francisco.

PFEFFER, J. (1972). Size and Composition of Corporate Boards of Directors: The Organization and its

Environment. Administrative Science Quarterly , 17, 2, 218-228.

RAJAN, R.; ZINGALES, L. (2004). Salvando o Capitalismos dos Capitalistas.

1ª Ed. Elsevier. Rio de Janeiro, 393 p.

SCHOORMAN, F.; BAZERMAN, M.; ATKIN, R. (1981). Interlocking directorates: A Strategy for reducing environmental uncertainty. Academy of Management Review , 6, 243-251.

SELZNICK, P. (1957). Leadership in Administration . New York: Harper e Row.

June 24-26, 2009


28


SHIVDASANI, A. YERMACK, D. (1999) CEO Involvement in the Selection of New Board Members:

An Empirical Analysis. The Journal of Finance.

. Vol. 54, N. 5, pp. 1829-1853(25)

STOKMAN, F., VAN DER KNOOP, J. WASSEUR, F. (1988) Interlocks in the Netherlands: Stability and careers in the period 1960–1980. Social Networks . Vol. 10; pp. 183-208.

TOBIN, J. (1958). Liquidity preferences as behavior towards risk. Review of Economic Studies , 35,

65-86.

WILLIAMSON, O. (1999). The Mechanisms of Governance , Oxford University Press. 439p.

ZAJAC, E.

(1988).

Interlocking Directorates as an Interorganizational Strategy: A Test of Critical

Assumptions.

The Academy of Management Journal , 31, 2, 428-438.

ZINGALES, L. (2000). In search of new foundations. Journal of Finance , 55, 1623-1653.

June 24-26, 2009


29


Table 1: Composition of the Board of Directors

This table shows descriptive statistics for the composition of Brazilian boards in 2003 and 2005. Outside directors are defined as those who only participate in the board. The outsiders’ percentage is defined by the ratio of outsiders to total board members. The number of firm-level observations is presented above each year.

Year

Observations

2001 n= 345

2003 n = 319

2005 n = 320

Category

Median

Mean

Board

Size

Outsiders percentage

Board

Size

6

6.4

(2.88)


81%

Board

Size

6

6.4

(2.97)


80% 6

6.5

(standard deviation) (3.39)

80%

81%

(84%)

80%

(80%)

81%

(81%)

Minimum

10

3

3

0%

0%

3

3

33%

67%

3

3

33%

67%

30

70

5

7

80%

79%

5

7

71%

89%

5

8

71%

92%

90 11 79% 10 100% 11 100%

Table 2: Directors’ relationships and interlocking directorates

This table displays different forms of interlocking in the sample in 2003 and 2005. The percentages showed in the right side are the proportions of firms in each interlocking category. Total of 345 firm-level observations in

2001, 319 in 2003 and 320 in 2005.

2001 2003 2005 Category

Companies with unified functions of CEO and chairman of the board

Companies with board interlocking

Companies whose CEO holds seat in another company

Companies whose chairman of the board holds seat in another company

Companies whose chairman of the board, who is also CEO, holds seat in another company

Companies in which half or more of the outside directors work in

2 or more companies

Companies in which half or more of the outside directors work in

3 or more companies

33%

65.3%

30.7%

46,7%

38%

38%

18%

31.3%

74%

31%

68.4%

32.6% 35.4%

47.6% 48.4%

42%

40%

30%

44%

39.20%

26%

June 24-26, 2009


30


Table 3: Mean number of linked firms to a given company trough shared directors

This table presents, for 2003 and 2005, the number of interlocks in our sample. Total of 345 firm-level observations in 2001, 319 in 2003 and 320 in 2005.

2001 Year

Mean

Median

Standard Deviation

3.0

4.9

2003

5.5

2

7.9

2005

4.8

3

6

Minimum

10

30

70

6.4

0

0

0

6

0

0

0

6

0

0

0

6

90 14 15 22

Maximum 32 39 30

Table 4: Frequency of board interlocking in Brazil

This table shows the dispersion of interlocking directorships among directors and officers (here called

“Administrators”). The left column of the table shows the number of directorships held by a given administrator

(from a minimum of one to a maximum of eight directorships). The other columns show how many professionals hold that number of positions and the total number of positions held in each category (e.g. in 2003, 2 administrators held 7 directorships, totaling 14 positions held by only 2 people). Total of 345 firm-level observations in 2001, 319 in 2003 and 320 in 2005.

Number of positions held by an

Administrator

8

7

6

3

2

5

4

1

Total

Administrators

2001

Total

Positions

- -

2

3

14

18

8

24

74

207

1443

1761

40

96

222

414

1443

2247

2003

AdminisTotal trators Positions

1 8

2

2

14

12

22

47

107

232

2275

2688

110

188

321

464

2275

3392

Adminis-

2005

Total trators Positions

- -

-

1

-

6

21

38

106

272

2301

2823

105

152

318

544

2031

3156

June 24-26, 2009


31


Table 5: Probit results assessing the probability of occurrence of board interlocking

This table presents results for a probit regression where the dependent variable is a dummy for the existence of interlocking. Total of 153 firm-level observations for 2003 and 153 for 2005. All t-statistic tests are double-tailed and *, ** and *** denote significance at the 10%, 5% and 1% levels, respectively.

Probit Regression

Dependent variable:

INTERLOCK

Total of

Observations:

Pseudo-R 2

306

0.42

Coefficient p-value Explanatory Variables

Board size

Independence (percentage of outsiders in the board)

CEO and Chairman are the same person

Concentration of ownership by the shareholder controlling block

Financial leverage – ln(DEBT)

Tobin’s Q

Return on assets

Firm size – ln(assets)

Participation in BOVESPA’s Level 2 or New Market

Year dummy

Constant

0.14

0.85

-0.02

-0.99

-0.01

-0.05

0.35

0.52

0.89

-0.07

-8.15

0.02

***

0.13

0.95

0.10

0.00

0.05

0.00

*

0.87

0.46

0.36

**

0.79

***

***

June 24-26, 2009


32


Table 6: Summary results for fixed effects and random effects estimations

This table shows the results of Fixed Effects and Random Effects estimations of four different models of the effect of board interlocking on firm value. The dependent variable is Tobin’s Q. In each model we use a different proxy for interlocking. The estimated coefficients are shown below (p-values in parentheses). In all regressions, the hypotheses of homoscedastic and normally distributed errors cannot be rejected at conventional levels. Total of 456 observations.

Fixed Effects Estimations Random Effects Estimations

Explanatory variables

Average number of seats occupied by the directors

CEO or Chairman of the Board participate in another company

Number of interlocks in the company

Half or more of the outside directors hold 3 or more positions

Board size

Board independence – percentage of outsiders


Percentage of ownership of the controlling block

Company size (logarithm of assets)

Return on Assets

Gross debt divided by total assets (financial leverage)

Constant

R²

Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4

-0.11

(0.45) -

-0.17

(0.22)

- -

-0.07

(0.29) -

-0.27

(0.02)

- -

- -

-0.03

(0.32)

- - -

-0.01

(0.41)

-

- - -

-0.30

- - -

-0.30

-

0.08

-

0.07

-

0.13

(0.07)

0.08

-

0.04

-

0.04

-

0.05

(0.02)

0.04

(0.37) (0.35) (0.30) (0.32) (0.22) (0.22) (0.22) (0.25)

0.95 1.00 1.12 0.73 -0.34 -0.16 -0.33 -0.43

(0.37) (0.32) (0.34) (0.44) (0.42) (0.71) (0.42) (0.32)

0.03 0.01 0.03 0.05 -0.23 -0.21 -0.23 -0.23

(0.92) (0.97) (0.91) (0.84) (0.27) (0.30) (0.28) (0.27)

0.65 0.60 0.67 0.59 -0.61 -0.62 -0.60 -0.65

(0.61) (0.63) (0.61) (0.63) (0.43) (0.43) (0.43) (0.41)

2.05 2.03 2.05 2.06 -0.13 -0.12 -0.13 -0.12

(0.12) (0.13) (0.12) (0.12) (0.35) (0.36) (0.34) (0.37)

1.45 1.54 1.42 1.50 2.00 2.12 2.01 2.04

(0.17) (0.14) (0.19) (0.15) (0.01) (0.01) (0.01) (0.01)

0.12 0.18 0.13 0.14 1.59 1.61 1.60 1.58

(0.86) (0.78) (0.84) (0.83) (0.00) (0.00) (0.00) (0.00)

-3.06 -3.04 -3.11 -3.06 2.82 2.63 2.74 2.85

0.24 0.24 0.25 0.25 0.55 0.55 0.55 0.56

June 24-26, 2009


33


Table 7 : 2SLS (Instrumental Variables) Estimations and GMM Estimations .

This table shows the results of regressions of corporate value, measured by Tobin’s Q, on different proxies for board interlocking. The estimated coefficients are shown below

(p-values in parentheses). In all regressions, the hypotheses of homoscedastic and normally distributed errors cannot be rejected at conventional levels. Total of 456 observations.

2SLS Instrumental Variables Estimation GMM Estimations

Explanatory Variables

Average number of seats occupied by the directors

CEO or Chairman of the Board participate in another company

Number of interlocks in the company

Half or more of the outside directors hold 3 or more positions

Board size

Board independence – percentage of outsiders


Percentage of ownership of the controlling block

Company size (logarithm of assets)

Return on Assets

Gross debt divided by total assets (financial leverage)

Constant

R²

Model 1 Model 2 Model 3 Model 4 Model 1 Model 2 Model 3 Model 4

-0.22

(0.04) - -

-

-

-

0.06

(0.12)

-0.73

(0.15)

-0.38

(0.07)

-0.73

(0.27)

-0.14

(0.26)

1.46

(0.11)

1.61

(0.00)

2.99

(0.51)

-

-0.37

(0.18)

-

-

0.05

(0.13)

-0.56

(0.25)

-0.29

(0.13)

-0.72

(0.27)

-0.14

(0.25)

1.68

(0.05)

1.64

(0.00)

2.81

(0.22)

-

-0.04

(0.05)

-

0.09

(0.07)

-0.64

(0.18)

-0.44

(0.06)

-0.64

(0.30)

-0.15

(0.23)

1.39

(0.14)

1.62

(0.00)

2.60

(0.23)

-

-

-0.57

(0.02)

0.05

(0.15)

-0.77

(0.15)

-0.40

(0.07)

-0.72

(0.27)

-0.15

(0.24)

1.57

(0.08)

1.61

(0.00)

2.78

(0.51)

-

-0.50

(0.11)

-

-

0.03

(0.13)

-0.38

(0.27)

-0.25

(0.10)

-0.54

(0.29)

-0.15

(0.21)

1.29

(0.24)

1.72

(0.00)

3.09

(0.52)

-0.26

(0.04)

-

-

-

0.04

(0.10)

-0.63

(0.14)

-0.37

(0.05)

-0.56

(0.27)

-0.14

(0.23)

1.00

(0.40)

1.33

(0.00)

3.56

(0.49)

-

-

-

-0.46

(0.01)

0.03

(0.18)

-0.79

(0.10)

-0.31

(0.07)

-0.62

(0.24)

-0.13

(0.23)

1.19

(0.30)

1.67

(0.00)

3.11

(0.50)

-

-

-0.05

(0.05)

0.08

(0.05)

-0.52

(0.19)

-0.44

(0.05)

-0.43

(0.34)

0.86

(0.49)

1.70

(0.00)

3.48

(0.47)

June 24-26, 2009


34


Table 8: Tests of quadratic effects of board interlocking.

This table shows the results of Pooled OLS estimations assessing the quadratic effect of the interlocking variables and allowing us to detect the optimal level of interlocking.

As dependent variables, we use four proxys for firm value and profitability: Tobin’s Q, PVB ratio, ROA and ROE. The main explanatory variables, used in separate regressions, are: LINK : total number of company connections made by its administrators; and PERCENTBUSY : ratio of “busy” directors to total board members. Control variables (essentially the same shown in Equation 2) were included in all regressions but are omitted from the table to save space. Estimated coefficients are shown below (pvalues in parentheses).

Explanatory variables Tobin’s Q PBV ROA ROE ln(LINK) ln(LINK)²

-0.16

(0.52)

-0.09

(0.67)

0.063

(0.009)

-0.018

0.019

(0.80)

PERCENTBUSY

PERCENTBUSY²

-0.02

(0.78)

-1.04

(0.27)

0.62

(0.50)

-0.02

(0.70)

-1.87

(0.03)

1.86

(0.01)

(0.011)

0.181

(0.004)

-0.232

(0.001)

-0.006

(0.80)

0.207

(0.60)

-0.208

(0.63)

June 24-26, 2009


35


Chart 1: Illustrative mapping of a board interlocking in 2005

Embraer SA

Gerdau SA

Mr. Boris

Tabacoff

Mr. Oscar

Bernardes Neto

Mr. Jorge Gerdau

Johannpeter

Suzano Papel e Celulose SA

Mr. Oscar

Bernardes

Neto

Alpargatas SA

Petrobras SA

Mr. Marco

Bologna

Mr. Boris

Tabacoff

TAM Transportes

Aereos SA

Klabin SA

Mr. Roger

Wright

Mr. Henri

Reichstul

Gradiente SA

Suzano

Petroqu í mica SA

Vivo Participa

Gol Linhas A é

ç ões SA

Mr. Luiz

Kauffman reas SA

Mr. Roger

Agnelli

Caemi SA

Stoliar

Figure 1: Return on assets (ROA) as a function of the number of interlocks per company.

Mr. Roger

Agnelli

Doce SA

Mr. Gabriel

Cia. Vale do Rio

Usiminas SA

June 24-26, 2009


36


1 Previous works in this field also employ the terms “ interlocking directorate”, “board overlap”, and “ board ties ” to describe the same phenomenon.

2 Dooley (1969) reports that, in 1965, only 3165 professionals occupied the 4007 positions on the boards of the

250 largest American companies. Although most of these board members occupied only one position, some professionals were active in more than nine companies.

3 Fich and Shivdasani (2006) observe that companies whose boards included a majority of professionals interlocked with other companies had, ceteris paribus , a price-to-book value approximately 4.2% lower, in comparison with other companies without this practice.

4 According to the definition by Fich and Shivdasani (2006).

5 Although it is also frequent among internal board members, Mizruchi (1996) highlights: “ Most interlocks are created by a firm’s outside directors”

6 Lipton and Lorsch (1992) observe that the most frequent problem board members face is the lack of time to perform their activities. Shivdasani and Yermack (1999) argue that the benefits of participating in other boards can be seen as a non-linear function, so that the participation in many boards can excessively reduce the time the board members have available to adequately carry out their monitoring responsibilities.

7 It should be highlighted that Fich (2005) also observes positive abnormal returns for the stock of companies that accept the CEO of another company as a board members. The author suggests that the American market perceives CEOs as very valuable for the company’s human capital, due to their management skills and competencies. This result is in line with the thesis that board interlocking is mainly ruled by the market forces that shape competition for human capital.

8 The concept of reputation is loaded with subjectivity. International research on the theme (e.g. Fich and White,

2001) has used the number of directors and executives in a given board as a proxy for the company’s reputation, due to these professionals’ personal reputation. In view of the lack of data in Brazil, we use two proxies for reputation: the company’s participation in the most advanced segments of BOVESPA corporate governance

(Level 2 and New Market) and the natural logarithm of the company’s assets, assuming that larger companies are more stable and, hence, have a more solid reputation history.

9 See Williamson (1999) for further details.

10 Fich (2005) highlights that qualified human capital can grant companies better competitive margins. Zingales

(2000) and Rajan and Zingales (2004) note the growing importance of human capital in corporate value measurement.

11 Tobin’s Q is a commonly used indicator of corporate value as, according to Morck, Shleifer and Vishny

(1988), it may capture the added value of intangible factors, such as the quality of the company’s corporate governance.

12 An exception to this rule occurs, according to the authors, when the chairman of the board of a company i is invited to participate in company j – in this case, exceptionally, an increase in the firm value of j is observed.

13 However, the authors remark that they cannot eliminate the endogeneity from their empirical model: companies with higher Tobin’s Q may be more attractive for other professionals to participate as board members; but it is also possible that these professionals’ participation makes the companies achieve superior results.

14 The liquidity index is computed by the Economática® database system. It is calculated based on the following formula:

LIQ



100 *



 p q



 * n

N

* v

V where: LIQ – stock liquidity index; p – number of days in which at least one deal was completed with the stock in a given period; P – total number of days in the period; n – number of deals with the stock in the period; N – number of deals with all stocks listed in the period; v – monetary value of the deals with the stock in the period;

V – monetary value of the deals with all stocks in the period.

15 In order to avoid errors and omissions, we checked each person’s official identity number, trying to capture possible cases of homonyms and guaranteeing the correct assessment of how many functions the professional occupied.

16 The criterion for classifying insider and outsider directors followed CVM’s methodology, according to which the directors who are also executive officers are classified as insiders, while those who solely work in the board of directors are classified as outsiders. Directors classified as substitutes (alternates) were excluded from the database.

17 The criterion to define the “stock control block” strictly followed the information in the CVM Annual Reports, in which stockholders with more than 5% of stocks are identified and designated as controllers or not, according to the information provided by the companies themselves.

June 24-26, 2009


37


18 The corporate governance listing segments (Level 1, Level 2 and New Market) of Bovespa are trading segments designed for companies that intend to adopt better corporate governance practices by complying with a series of requirements compatible with each level. Only companies participating in Level 2 and the New Market were selected, since Level 1 firms must comply with only minor requirements, essentially related to stock liquidity and information transparency. Level 2 and New Market are segments that actually require better corporate governance practices and grant extended rights to minority stockholders.

19 Again, it’s important to emphasize that, since we only look at publicly traded companied listed on Bovespa with some stock liquidity, the level of board interlocking may be understated, since some directors may serve as board members of other companies that are not included in the sample, such as private companies or public companies not traded in the stock exchange.

20 In the vast majority of cases, these situations occur in companies that belong to the same economic group.

21 Tobin’s Q was estimated using Chung and Pruitt’s approach (1994):

Q



MVCS



MVPS



D ; where: MVCS is

TA the market value of common stocks; MVPS is the market value of preferred stocks; D is the book value of shortterm debt plus long-term debt less current assets and TA stands for the book value of the company’s total assets.

22 This positive (‘U’ shaped) quadratic relation would indicate that low or extreme levels of interlocking are positive for company value, while intermediate levels are harmful.

23 Non and Franses (2007) observe a positive quadratic relation, though with low statistical significance, for the

PBV and price-earnings ratios. However, the authors do not present theoretical justifications for this evidence.

For the profitability variables ROA and ROE, an inverted quadratic relation was found for the interlocking variable (mean number of seats occupied by the board members).

24 The reputation of the firm was proxied by the participation of the firm in BOVESPA’s differentiated corporate governance levels.

June 24-26, 2009


38

Board Interlocking In Brazil: Directors’ Participation In Multiple Companies And Its Effect On Firm Value

Board Interlocking in Brazil: Directors’ participation in multiple companies and its effect on firm value

Rafael Liza Santos

Alexandre Di Miceli da Silveira

Lucas Ayres B. de C. Barros

Board Interlocking in Brazil: Directors’ participation in multiple companies and its effect on firm value

Abstract

JEL classification code:

Related documents

Products

Support

Board Interlocking In Brazil: Directors’ Participation In Multiple Companies And Its Effect On Firm Value

Board Interlocking in Brazil: Directors’ participation in multiple companies and its effect on firm value

Rafael Liza Santos

Alexandre Di Miceli da Silveira

Lucas Ayres B. de C. Barros

Board Interlocking in Brazil: Directors’ participation in multiple companies and its effect on firm value

Abstract

JEL classification code:

Related documents

Add this document to collection(s)

Add this document to saved

Suggest us how to improve StudyLib